Traders need open minds. Many of us believe trading systems are the only way to enjoy long-term success in the markets. Discretionary traders like to take a more active hand in trading. System traders can be both fundamental and technical, and though discretionary traders aren’t mechanical, they often make trading decisions based on the same technical analysis as system guys.
The differences between these approaches are important to practitioners of each. For example, prudent discretionary traders will backtest parts of their trading methods, and system traders should understand some key aspects of discretionary trading to develop reliable mechanical trading systems.
These differences will help in examining how today’s markets demand a new approach in systematic trading.
MAN VS. MACHINE
Great discretionary traders are born. It is a God-given gift, such as the ability to throw a 100-mile per hour fastball. Few people are capable of succeeding as discretionary traders. On the other hand, system traders can be created.
System traders come in two types. Those who purchase commercial systems need to do their own research and learn to trust their system with the amount of risk they are assuming. Learning to trust a system requires you to study every trade to determine if it can be followed when difficulties occur.
The second kind of system trader creates his own trading system. If you build your own systems, you need to understand why a system should work. Establishing a viable premise is the first step professional system developers take.
While most successful discretionary traders have a seemingly inherent understanding of the market, they use more than gut feel in their decision making. Many of the best discretionary traders are systematic in their approaches, but not mechanical. An example is that they might draw trendlines using price or indicators. How the trendlines are drawn is the subjective part, but how they are traded could be mechanical. (see “Systematic execution.”)
The chart shows crude oil prices from April through May 2006 with the moving average convergence-divergence (MACD) plotted in the bottom panel. The MACD was making lower highs while crude oil was retesting its highs. This divergence led to a $5,000 move in crude in less than two weeks.
Discretionary traders also use systems as indicators or entries. The art is in the exits and deciding when to employ different trading methodologies.
The ability to perform at a high level every day is a gift. The difference is like that between a minor league baseball player and a major league one. Minor league players often come up to the majors and play well for a while, but they nearly always return to the minors because they lack consistency.
The problem for both traders and baseball players is handling pressure. There are baseball players who hit well except in clutch situations, and there are many traders who can string together a number of small profitable trades. But it’s when the bases are loaded or volatility is extreme, that great baseball players and great traders are made.
That is where system trading can help. If you can develop and trade mechanical systems, you only have to follow the system. If you lack the discipline to trade your system, there are a number of brokers who specialize in trading these for you. Bottom line: If you can build a quality trading system, then you can make money in the markets.
THE RIGHT FIT
For you to reliably follow a trading system, it needs to match your risk tolerance and reward expectations. If you find it emotionally difficult to deal with 40% winning trades, you will find it difficult to trade a trend-following method. Also, if you cannot deal with trades that last months, either winners or losers, then trend-following systems are not for you. However, if you want to manage the trading of the system yourself but you hold a 9-to-5 job, then don’t even think about an intraday trading system.
The markets you trade also should depend on your personality. If you are risk-averse, you do not want to attempt to trade markets that are highly volatile like the energy complex is now, or like the S&P 500 was during the late 1990s and early 2000s. Large changes in daily equity in either direction might cause you not to follow the rules of your system.
Before selecting a system to trade you must ask three questions: Did you do thorough homework on the system? Do you trust the system? And, will you follow the system? Be honest with yourself when you answer them. If your answer is “no” to any question, you should look elsewhere or, if you must trade this system then consider taking your business to a system-assist broker. It is their job to just follow the rules, and most of them are pretty good at it.
INTRADAY DATA ISSUES
Developing intraday trading systems is more difficult than end-of-day systems because intraday data is more expensive and a longer data history is needed to develop a reliable system. The problem is if the data history does not include enough time, the system will not be tested on certain types of markets. The system may have seen nothing except the market conditions that currently exist. This could lead to major problems as market conditions change.
A good standard is to use at least three, preferably five, years of intraday data to develop a reliable intraday trading system. You may discover markets the system doesn’t perform well in, which is the point. You also will understand under what condition it does work, which allows you to adapt if performance degrades.
Another problem is that electronic markets sometimes have different market noise patterns than pit-traded markets. To overcome this, many system developers will develop and run their trading systems on the pit-traded markets when they’re available, but execute the trades in the electronic markets. This is also true for stops, which often are hit prematurely on electronic markets but not on the pit-traded markets. To see this for yourself, examine the time-and-sales data of the E-mini and S&P 500 contracts. Every few months, you will see a weird E-mini trade taken down because it was not representative of market price action. If you don’t believe that “one weird trade” matters, just ask the trader who was on the wrong side of it when it happened.In cases where it is an option, system performance needs to be evaluated on both markets so you can judge the effect on system performance.
Trading today is not just about the U. S. markets. Many of today’s U. S. markets are less liquid than they have been in the recent past. Through the years, some of these markets have done well for trend-following systems but traders have started to look for different markets to trade because of liquidity issues.
Many overseas markets trend well, including the German bund and the Japanese government bond (JGB). The problem with developing systems for trading these markets combined with U.S. markets is that these markets are quoted in their respective native currency. The bund is quoted in euro, and the JGB is quoted in yen. So, you cannot just include them in your system backtest run and use the testing results.
The results need to be converted from the native currency to the account currency, which in most cases is dollars. Unfortunately, most backtesting platforms do not have this capability built in. Three trading platforms that do this are TradersStudio, Trading Blox and Mechanica.
To the trader, trading overseas markets is just like trading the U. S. markets. Overseas markets provide a new source of liquidity and they often trend well. In case you were wondering, when you trade these markets with your broker, the currency conversion is automatically done in your account.
The first step in developing trading systems is to decide what you want to trade. Then, if you do not have a system development platform, you can get one based upon your perceived needs. If you want to trade a basket of commodities or stocks, you need a platform with portfolio analysis capabilities. If you are trading a small account, then money management might not be as important in your platform decision.
If you are looking at trading non-U.S. markets you should have currency conversion capabilities. If you trade forex, you need currency conversion and support for profit/loss adjustment based on the interest rate differential. You will also want a rich programming language so your ideas are not limited by the platform you are developing systems in. After that, select a clean and reliable source of data and determine how much data you need.
Once you have selected markets, you need to decide on your trading time frame. From there, you need to decide which methodologies you will rely on for your analysis.
The next installment will look at both adopting existing methodologies and building your own.
Murray A. Ruggiero Jr. is a consultant in East Haven, Conn. His firm, Ruggiero Associates develops market-timing systems. He is editor-in-chief of Inside Advantage Gold Club (www.iagoldclub.com) and is the author of Cybernetic Trading Strategies (John Wiley & Sons). E-mail: email@example.com.